Reservoir characterization is aimed at identifying hydrocarbon bearing reservoirs, delineating them and subsequently, determining the distribution of relevant physical properties such as lithology, porosity, permeability, water saturation and pore pressure, which will make for an easy determination of the reservoir’s economic potential [1] . Prospect definition requires more than just mapping of geologic structures that can bear hydrocarbon. It entails a more quantitative evaluation of both the static and dynamic properties of the specific reservoir.

Recent interests in reservoir characterization focus on optimizing the clarity of the geological and reservoir simulation models of prospects, by accurately defining both the static and dynamic properties of reservoirs, through the integration of petrophysics in the interpretation of seismic data. The use of inversion algorithms based on the approximations of Zoeppritz’s equations has been studied by various researchers [2] [3] , for the inversion of prestack seismic data into acoustic and shear impedance volumes. Their results demonstrated the use of seismically derived attributes such as acoustic impedance, lambda-rho, Poisson impedance and Murho as effective tools for lithology and fluid prediction in a hydrocarbon reservoir.

During the production life of a reservoir, changes in fluid saturation, pressure and temperature induce changes in the reservoir’s density and compressibility properties. These changes undoubtedly, will affect the seismic response of reservoir rocks which could be investigated over a time lapse using (4-D) seismic data. This serves as a veritable tool to monitor and update reservoir models for optimized hydrocarbon recovery.

The study area lies to the NE of Niger delta in southeastern Nigeria, situated on the continental margin of the gulf of Guinea (Figure 1). The present work attempts to enhance hydrocarbon prospect definition in the field by integrating well and 4-D seismic data sets for effective monitoring of the in-situ reservoir properties over the producing regions and map possible by-passed hydrocarbon resources in the field.

The Niger delta is a sedimentary basin formed as a complex regressive off lap of clastic sediments and contains only one identifiable petroleum system referred to as the tertiary Niger delta (Akata-Agbada) petroleum system. This is characterized by structural and stratigraphic hydrocarbon plays overlain by the most recent Benin Formation [4] .

Preliminary analysis of the 4-D seismic data shows that the producing reservoir is situated on the crest of a multi-fractured rollover anticlinal structure (Figure 2), bounded by a major growth fault on the North-East of the field. The significant petroleum reservoir sands in the field consist mainly of middle Miocene deltaic sandstones, which are poorly consolidated, with high effective porosities and permeability’s [5] . This makes the field ideally suited for a time-lapse analysis, as effects due to fluid saturation change could be readily discernible.

Logs from producing wells (A, B, and C) and 4-D seismic data from the field were used in this study. The well logs comprising Density, Gamma ray, Sonic and Resistivity were de-spiked using a median filter, after corrections for mud filtrate invasions, well bore wash outs and missing data. Gamma ray and Resistivity logs were

used in delineating HD2000 hydrocarbon reservoir based on low gamma ray counts and high electrical resistivity values (Figure 3). V_pV_s ratio, Poisson’s ratio, Lambdarho, Murho, acoustic and shear impedance logs were derived from appropriate rock physics empirical models [6] - [9] . These rock attributes were analyzed over the reservoir interval in 3-D cross plot space and by Gassmann’s fluid substitution modeling for varying oil and brine saturations. Generally, common lithologic units and fluid types tend to form distinct clusters in cross plot space, and this helps in making a straight forward interpretation of probable lithology and pore fill saturant.

The original oil-water contact (OOWC) for HD2000 reservoir was about 5900 ftss while as at the year 2000, the oil water contact (OWC) was at about 5856 ftss. Average pressure and gas-oil ratio were 2532 psi and 298 scf/stb respectively, for well B. Well information’s were not available for wells (A and C) in this study.

The 4-D seismic data were reprocessed in parallel to ensure no extraneous effects due to acquisition footprint and processing shows up during 4-D analysis. Also, a cross correlation test was carried out between the vintages to ensure repeatability. Minimum correlation coefficient of 0.61 and maximum of 0.85 were obtained around and beyond the producing wells (Figure 4). This is an indication that the repeatability between the two vintages is good enough to allow for effective visualization of production induced effects in the field.

Subsequently, the seismic data from the two vintages were independently inverted into acoustic impedance volumes through a model based inversion scheme. These inverted impedances were used to generate the Lambda-mu-rho attributes for saturated media [5] [8] [10] , to quantitatively characterize the reservoir in the field based on the results of well log cross plot analysis. The generation of Lambda-mu-rho attributes was done using a multi-attribute neural network algorithm. This algorithm works by identifying all possible linear and non- linear relationships that may exist between the desired log attribute and the several available seismic attributes at the well location. Based on these relationships, networks are trained to predict the log attribute over the seismic volume away from the well position in a least-squares sense, after validation.

The petrophysical well log analysis in cross plot domain and fluid sensitivity modeling using Gassmann’s fluid substitution, revealed that low ρ, λρ and I_p, and high R are associated with gas and oil sands, while medium to high ρ, λρ and I_p, and low R values are associated with brine sands and shale, respectively. Results also show that λρ attribute is the most robust in discriminating for reservoir pore fill than the other investigated properties for all saturations. ρ and I_p properties are only sensitive to saturations above 75% oil saturation.

The sensitivity of ρ and I_p properties to oil and brine sands is dependent on saturation. At low oil saturation, ρ and I_p do not distinctly distinguish oil from brine sands, but becomes more responsive with increasing oil saturation. This is attributed to the comparable densities of these reservoir saturants, such that increasing oil saturation will mask the effect of brine density on the responses of ρ and I_p properties in the reservoir.

Reservoir sands plotted as low μρ and I_p, and high R clusters than shale. This low μρ for oil sands is in variance with the high μρ associated with hydrocarbon sands from petrophysical studies [7] . Low μρ value suggests that the reservoir sands are probably unconsolidated and might be associated with minor shale intrusions in the investigated field.

The horizon attribute slices delineated the producing well locations as low ρ, λρ, I_p and water saturation zones in the base compared to the monitor slice, which is consistent with the results of the log-based analysis. The difference slice show that well B location is delineated by moderate ρ, λρ, I_p and water saturation values, while wells A and C lies on higher values. These results are obvious indications that well B will soon water out, while wells A and C have watered out. This effect is attributed to hydrocarbon withdrawal and replacement by brine associated with pore pressure depletion in the reservoir.

The region north of well B is associated with no observable 4-D change and is thus, considered to be a probable by-passed hydrocarbon zone. The presence of NE and SW faults bordering this region supports this pay proposal, as the faults are adequate seal structures for hydrocarbon accumulation in the field. Other mapped zones (A and B) which do not show convincing prospects for hydrocarbon accumulations from petrophysical analysis, could be further investigated for probable by-passed resources.

The horizon attribute slices show that low ρ, λρ and I_p associated with hydrocarbon charged sands are consistent with the result of the log-based petrophysical analysis. These rock attributes were observed to have increased in magnitude on the monitor slices, indicating pore fill saturation changes and pore pressure depletion. Wells A and C have watered out, while well B will soon water out.

The region north of well B within the producing zone however, shows no observable 4-D effect and could be considered as a bypassed hydrocarbon prospect. The presence of faults bordering this zone supports this pay proposal. Other mapped zones (A and B), which do not show convincing 4-D effects should be subjected to further petrophysical investigation.

We express our appreciation to the HAMPSON-RUSSELL Corporation and the SHELL PETROLEUM DEVELOPMENT COMPANY Port Harcourt, for their technical support and permission to use data for this study, without which we would not have succeeded.

Tamunonengiyeofori Dagogo,Chukwuemeka Ngozi Ehirim,Joseph Onukansi Ebeniro, (2016) Enhanced Prospect Definition Using Well and 4D Seismic Data in a Niger Delta Field. International Journal of Geosciences,07,977-990. doi: 10.4236/ijg.2016.78074